Fast 1xN fiber-optic switch

BACKGROUND

Recently, fiber optics networks have been used instead of electrical systems to provide significantly higher bandwidth. However, some of these fiber optics networks employ electrical switches to direct communication signals from various input ports to one or more output ports. In these systems, the optical signal is converted to an electrical signal prior to the switching process. The electrical switch redirects the transmission of the electrical signal which is then converted back to an optical signal before being transmitted to the output port.

In all-optical networks, optical switches are used to direct a signal from one port or channel to another port or channel. For example, some have proposed using mechanical optical switches, such as micromachined mechanically actuated mirrors, or liquid crystal based switches, to redirect the light signal. As all-optical networks increase in complexity, the speed at which the optical switching process occurs becomes more important. But because the switching times of mechanical optical switches are in the order of milliseconds, and liquid crystal based switches offer only slightly higher speeds, their use is limited.

All-optical switches based on electro-optic (E-O) effects of certain solid-state materials offer significantly higher switching times, for example, in the order of microseconds to nanoseconds. Some proposed systems use E-O switches based on polarization conversion, while others use phase modulation.

It has also been proposed to use optical switches based on E-O beam deflection effects to provide high-speed, multi-channel switching. Typically, however, the light being transmitted through an ordinary optical fiber is randomly polarized while the E-O effect is polarization dependent. To make the fiber optic switching process polarization independent, some systems include an optical device that converts the incoming light signal into two linearly polarized light beams having the same polarization direction. An E-O based beam deflector is used to deflect these two beams which are then combined by a second optical device into a single output beam. In these system, each of the optical devices is typically formed from a pair of prisms which are expensive and difficult to fabricate. Further, these prisms typically cannot be made in a compact size.

SUMMARY

Birefringent crystals can be used as compact devices for separating a randomly polarized beam into two beams with orthogonal polarization directions, or combining two such beams into a single beam based on “walk-off” effects. These devices alone, however, are unsuitable for use as beam deflectors. Moreover, the walk-off distance between the two beams that have been separated or that are to be combined varies with the angle of incidence with the walk-off device. Accordingly, these variations cause two incoming beams to recombine imperfectly, and hence induce a polarization dependent loss. Therefore, there is a need to use walk-off devices with a compensator that eliminates these variations to provide an optical device with low polarization dependent loss.

The present invention implements a high-speed 1×N optical switch with an Electro-Optic (E-O) deflector. The optical switch redirects or steers an incoming light signal to a number of output light ports in a polarization independent manner.

In one aspect of the invention, the optical switch includes a walk-off device, and a compensator which compensates for walk-off distance variations of two polarized beams, with orthogonal polarization directions, associated with the walk-off device.

In some embodiments, the walk-off device is a crystal with birefringence such as, for example, yttrium vanadate or rutile. The compensator can be an isotropic plate such as glass. In certain embodiments, the walk-off device is arranged to separate an incoming beam into the two polarized beams, while in other embodiments, the walk-off device is arranged to combine the two polarized beams into a single beam.

In another aspect of the invention, the optical switch includes a separator which separates the incoming light signal into a first polarized beam and a second polarized beam so that the polarization direction of the first beam is orthogonal to that of the second beam. A first converter rotates the polarization direction of one of the two beams to align the polarization direction of that beam with the polarization direction of the other beam. An electro-optic (E-O) deflector deflects the two beams and a compensator displaces one of the two deflected beams relative to the other beam to minimize polarization dependent losses. A second converter rotates the polarization direction of the displaced beam so that it is orthogonal to the polarization direction of the other beam. A combiner then combines the two beams into a single output light signal which is directed to the output port.

Embodiments of this aspect can include one or more of the following features. In some embodiments, the separator and the combiner are walk-off devices. The walk-off devices can be crystals with birefringence properties such as yttrium or rutile. The first and the second rotators can be Faraday rotators, quartz rotators, or half-wave plates, and the compensator can be an isotropic plate, such as glass.

In certain embodiments, the E-O deflector includes a prism having an index of refraction that changes as an electric field is applied across the prism. For example, the prism can be a crystal made from lithium tantalate, lithium niobate, or KTN. In other embodiments, the prism can be made from a ferroelectric ceramics. Alternatively, the prism can be a liquid crystal dispersed polymer.

Some switches have an E-O deflector having two or more prisms. In some embodiments, the prisms have linear E-O properties such that the deflection of the two beams produced by the prisms is additive when the electric field applied to each prism has the same polarity. In other embodiments, the prisms are configured in a domain inversion arrangement. In this arrangement, the deflection of the two beams produced by the prisms is additive when the electric field applied to a respective prism has a polarity that is opposite to that applied to an adjacent prism. In yet other embodiments, the prisms are made from materials with quadratic E-O properties.

The optical switch can include or can be coupled with a collimator which directs the incoming light signal to the separator. The collimator can include, for example, an input lens and an input port. The optical switch can also include an output collimating lens and an array of output ports. The input port and the output ports are typically optical fibers. The output ports can be coupled to a fiber pigtailed channel wave-guide array which directs the light signal from the output lens to the array of output ports.

Related aspects of the invention include a method of using the device as an ultrafast 1×N switch. The method includes separating the input light signal into a first polarized beam and a second polarized beam so that the polarization direction of the first beam is orthogonal to that of the second beam. The polarization direction of one of the two beams is rotated to align with that of the other beam. The two beams are then deflected, and subsequently one of the two deflected beams is displaced an appropriate distance to minimize polarization dependent losses. The polarization direction of the displaced beam is rotated so that its polarization direction is again orthogonal to that of the other beam. The two beams are then combined into a single output light signal.

Embodiments of the invention may have one or more of the following advantages. The optical switch of the present invention is able to quickly redirect a light signal from one output fiber to another in time periods in the order of nanoseconds. The optical switch is compact in size, reliable, and requires low power to operate. A particular advantage of the optical switch is that it can switch the direction of a light signal in a polarization independent manner.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and other objects, features and advantages of the invention will be apparent from the following more particular description of preferred embodiments of the invention, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention.

FIG. 1 is an illustration of an ultra-fast 1×3 optical switch in accordance with the invention.

FIG. 2A is an illustration of a light signal with a zero degree of angle of incidence passing through a walk-off crystal.

FIG. 2B is an illustration of a light signal with an angle of incidence greater than zero passing through a walk-off crystal.

FIG. 3A is an illustration of a walk-off crystal with a converter used for recombining two beams of polarized light having orthogonal polarization directions.

FIG. 3B is an illustration of the walk-off crystal and the converter of FIG. 3A combined with a compensator in accordance with the invention.

FIG. 3C is an illustration of the compensator plate of FIG. <b>3</b>B.

FIG. 3D is a graph showing the compensation effects of the compensator plate of FIG. <b>3</b>B.

FIG. <b>4</b>A(i) is an illustration of a multi-stage E-O deflector in accordance with the invention.

FIG. <b>4</b>A(ii) is a top view of the E-O deflector of FIG. <b>4</b>A(i).

FIG. 4B is an illustration of an alternative embodiment of the E-O deflector.

FIG. 4C is an illustration of yet another alternative embodiment of the E-O deflector.

FIG. 5A is an illustration of a fiber pigtailed channel wave-guide array.

FIG. 5B is a cross-sectional view of the wave-guide array taken along the line <b>5</b>B—<b>5</b>B of FIG. <b>5</b>A.

DETAILED DESCRIPTION OF THE INVENTION

A description of preferred embodiments of the invention follows.

Referring to FIG. 1, there is shown a 1×3 optical switch <b>10</b> which is provided with an E-O beam deflector <b>16</b> that deflects or redirects an incoming light signal from an input port such as an input fiber <b>12</b> to one of a set of output ports or output fibers <b>18</b>, <b>20</b>, and <b>22</b> in a polarization independent manner. As shown, an input collimating lens <b>13</b> can be used to collimate the light signal from the input fiber <b>12</b>. Typically, the input fiber <b>12</b> and the collimating lens <b>13</b> are formed together and pre-made as a collimator <b>14</b>. An output collimating lens <b>24</b> can also be used to collimate the light signal before it is transmitted to one of the output fibers <b>18</b>, <b>20</b>, and <b>22</b>. Although shown as a 1×3 switch, the optical switch <b>10</b> can easily be used as a 1×N switch, where N is the number of output ports that can be significantly larger than three.

The E-O deflector <b>16</b> includes a series of prisms <b>16</b><i>a</i>, <b>16</b><i>b</i>, and <b>16</b><i>c </i>made from, for example, a solid-state material. By applying an electric field across one or more of these prisms <b>16</b><i>a</i>, <b>16</b><i>b</i>, and <b>16</b><i>c</i>, the light signal is steered or redirected to one of the output ports.

In addition to the E-O deflector <b>16</b>, the optical switch <b>10</b> includes a separator <b>26</b> and a first polarization converter <b>28</b> positioned between the collimator <b>14</b> and the E-O deflector <b>16</b>, and a combiner <b>30</b> and a second polarization converter <b>32</b> positioned on the other side of the E-O deflector <b>16</b> between it and the output collimating lens <b>24</b>. The optical switch <b>10</b> also includes a compensator plate <b>34</b> placed between the E-O deflector <b>16</b> and the second polarization converter <b>32</b> to ensure that the switching process is polarization independent and occurs with minimal polarization dependent loss.

In use, the collimator <b>14</b> transmits input light signals <b>36</b> to the optical switch <b>10</b> as arbitrarily polarized light. The separator <b>26</b> converts the light signal <b>36</b> into two linearly polarized beams <b>38</b> and <b>40</b> having orthogonal polarization directions, as identified by the filled circle, and the thick bar, respectively. The first polarization converter <b>28</b>, placed in the path of one of the beams, for example, the beam <b>38</b>, rotates the polarization direction of that beam by 90° so that the two beams <b>38</b>′ and <b>40</b> have the same polarization direction.

If no electric field is applied across the prisms <b>16</b><i>a</i>, <b>16</b><i>b</i>, and <b>16</b><i>c</i>, the two beams <b>38</b>′ and <b>40</b> pass through the E-O beam deflector <b>16</b> as undeflected light beams <b>38</b>″ and <b>40</b>′. The second polarization converter <b>32</b> rotates the polarization direction of the beam <b>40</b>′ by 90° to a rotated beam <b>40</b>″ such that polarization directions of the beams <b>38</b>″ and <b>40</b>″ are now orthogonal to each other. The combiner <b>30</b> recombines these two beams into a single beam <b>36</b>′ that is again randomly polarized. The light beam <b>36</b>′ passes through the collimating lens <b>24</b> where the beam is focused before it enters the output fiber <b>18</b>.

To switch the light signal from output fiber <b>18</b> to one of the other output fibers <b>20</b>, <b>21</b>, and <b>22</b>, an electric field is applied across one or more of the prisms of the E-O deflector <b>16</b>. For example, to direct the light signal to the output fiber <b>20</b>, a sufficient voltage is applied to one of or a combination of the prisms <b>16</b><i>a</i>, <b>16</b><i>b</i>, and <b>16</b><i>b </i>to deflect the two light beams <b>38</b>′ and <b>40</b> downward as light beams <b>38</b>′″ and <b>40</b>′″. The converter <b>32</b> rotates the polarization direction of the beam <b>40</b>′″ to a rotated beam <b>40</b>″″ that has the same polarization direction as the beam <b>38</b>′″. The two beams <b>38</b>′″ and <b>40</b>″″ are combined by the combiner <b>30</b> to a randomly polarized beam <b>36</b>″ which is subsequently transmitted to the output fiber <b>20</b> via the collimating lens <b>24</b>.

The various components of the of the optical switch <b>10</b> will now be described in greater detail. Each of the polarization converters <b>28</b> and <b>32</b> in the illustrated embodiment is a half-wave plate. To rotate the polarization direction of an incoming beam 90°, the optical axis (or c-axis) of the half-wave plate is aligned to be perpendicular to the incoming beam propagation direction and either 45° or 135° to the polarization direction of the beam. Alternatively, the converters <b>28</b> and <b>32</b> can be Faraday rotators or quartz polarization rotators.

The separator <b>26</b> and the combiner <b>30</b> can be walk-off crystals made from any suitable birefringent crystal, such as, for example, yttrium vanadate or rutile. An illustration of how these crystal separate an incident light beam is shown FIG. <b>2</b>A. When a light signal <b>45</b> passes through a walk-off device <b>46</b> with an optical axis c at an angle of incidence of zero degrees, the orthogonal polarized components of the light signal separate into the two beams <b>47</b> and <b>48</b>. The separation distance between the two beams <b>47</b> and <b>48</b> is typically referred to as the walk-off distance, d_w. However, the walk-off distance, d_w, between the two beams is dependent on the incident angle, i, of the light beam <b>45</b> (FIG. <b>2</b>B). That is, the walk-off distance of the two beams <b>47</b> and <b>48</b> in FIG. 2A is slightly different than that of the beams in FIG. <b>2</b>B.

Note that although the walk-off distance referred to above is the distance between two beams that have been separated apart from a single incoming beam, the walk-off distance can also refer to the distance between two incoming beams that are combined together by the walk-off device into a single outgoing beam. In those circumstances, the angle, i, would refer to the deflection angle of the single outgoing beam.

The relationship between the walk-off distance, d_w, and the incident (deflection) angle, i, can be calculated by ray tracing methods. The results of such a calculation are shown below in Table 1 for an yttrium vanadate walk-off device with a thickness of approximately 10 mm and for a wavelength of about 1.55 μm.

<tables><table frame="none" colsep="0" rowsep="0"><tgroup align="left" colsep="0" rowsep="0" cols="1"><colspec colname="1" colwidth="217pt" align="center" /><thead><row><entry namest="1" nameend="1" rowsep="1">TABLE 1</entry></row><row><entry namest="1" nameend="1" align="center" rowsep="1" /></row><row><entry>Walk-off Distances for Various Incident/Deflection Angles</entry></row><row><entry namest="1" nameend="1" align="center" rowsep="1" /></row></thead><tbody valign="top"><row>&lt;entry /&gt;</row></tbody></tgroup><tgroup align="left" colsep="0" rowsep="0" cols="7"><colspec colname="1" colwidth="56pt" align="left" /><colspec colname="2" colwidth="21pt" align="char" char="." /><colspec colname="3" colwidth="28pt" align="char" char="." /><colspec colname="4" colwidth="28pt" align="char" char="." /><colspec colname="5" colwidth="28pt" align="char" char="." /><colspec colname="6" colwidth="28pt" align="char" char="." /><colspec colname="7" colwidth="28pt" align="char" char="." /><tbody valign="top"><row><entry>Incident/deflec-</entry><entry>0</entry><entry>5</entry><entry>10</entry><entry>15</entry><entry>20</entry><entry>25</entry></row><row><entry>tion angle,</entry></row><row><entry>i (degrees)</entry></row><row><entry>Walk-off</entry><entry>0.994</entry><entry>0.992</entry><entry>0.986</entry><entry>0.976</entry><entry>0.959</entry><entry>0.935</entry></row><row><entry>distance, d_w(mm)</entry></row><row><entry>Variation in d_w</entry><entry>0</entry><entry>−0.002</entry><entry>−0.008</entry><entry>−0.018</entry><entry>−0.035</entry><entry>−0.059</entry></row><row><entry>(mm)</entry></row><row><entry namest="1" nameend="7" align="center" rowsep="1" /></row></tbody></tgroup></table></tables>

As can be seen, the walk-off distance, d_w, for an angle of incidence/deflection, i, of zero degrees is 0.994, but note that the walk-off distance, d_w, decreases as the incident/deflection angle, i, increases, as indicated in the third row (variation in the walk-off distance). For the optical device <b>10</b> discussed above, this variation causes the two beams <b>38</b>′″ and <b>40</b>′″ to recombine imperfectly as two outgoing beams <b>36</b>″<i>a </i>and <b>36</b>″<i>b </i>thereby inducing a polarization dependent loss (FIG. <b>3</b>A). To reduce this effect, the compensating plate <b>34</b> is place in front of the combiner <b>30</b>, as shown in FIG. 3B, so that the two beams <b>38</b>′″ and <b>40</b>″″, recombine into the single beam <b>36</b>″ regardless of the deflection angle. Of course, the two beams <b>38</b>″ and <b>40</b>′ (FIG. 1) will recombine perfectly regardless whether the compensator plate <b>34</b> used, since the deflection angle of the outgoing beam <b>36</b>′ is zero. Further, the variation in the walk-off distance is not symmetrically the same for negative incident/deflection angles. However, an appropriately designed compensator plate can compensate for these negative incident/deflection angles as well. It should also be noted that the compensating plate <b>34</b> can be placed in the path of the beam <b>38</b>′″ to eliminate or minimize walk-off distance variations. Also, the compensating plate <b>34</b> can be place in the path of either of the two separated beams transmitted from the separator <b>26</b> (FIG. 1) to compensate for possible walk-off distance variations produced by the separator <b>26</b>.

The compensating plate <b>34</b> can be an isotropic plate with a thickness of T, as shown in FIG. <b>3</b>C. It is well known in the art that the displacement, d_c, of a light beam B varies with the angle of incidence, β. By choosing the plate <b>34</b> with an appropriate thickness and refractive index, it is possible to compensate for the walk-off distance variation produced by typical walk-off crystals. That is, by properly choosing the plate <b>34</b>, the displacement, d_c, of one of the two separated beams will compensate for the variation in d_wsuch that the two beams are combined by the combiner <b>30</b> into a single beam with minimal polarization dependent loss.

Referring now to FIG. 3D, there is shown an example of the compensation effects for a particular compensating plate <b>34</b> with a thickness, T, of about 0.3 mm and an index of refraction of 1.5, which is approximately the index of refraction of a glass plate. In this figure, the bottom curve <b>50</b> shows the walk-off distance variation of the two beams <b>38</b>′″ and <b>40</b>′″ (FIG. <b>3</b>B), the upper curve <b>52</b> shows the displacement, d_c, for example, of the beam <b>40</b>″″ (FIG. 3B) produced by the compensating plate <b>34</b>, and the middle curve <b>54</b> shows the compensated result which is essentially zero indicating that the polarization dependent loss is reduced nearly to zero.

Turning now to the E-O deflector <b>16</b>, there are shown in FIGS. 4A-4C three alternative embodiments of the multi-stage E-O deflector <b>16</b>. Again, each of these E-O deflectors <b>16</b> includes a set of prisms similar to those shown in FIG. <b>1</b>. In principle, the beam deflection properties of these deflectors are based on prisms having index of refractions that are varied by changing the electric field applied across these prisms. Thus, when no electric field is applied to a prism, the particular prism has an index of refraction of n₀. Applying an electric field across the prism induces a change in the index of refraction Δn so that the resulting index of refraction of the prism is n=n₀+Δn. Furthermore, a light beam traversing the prism is deviated by an amount depending upon the prism shape and its orientation, as well as the index of refraction induced by the electric field.

The E-O prism can be made from crystals having E-O properties such as lithium tantalate, lithium niobate, potasium tantalum niobium oxide (KTN). Other suitable E-O materials include ferroelectric complex oxides in the form of a transparent ceramic such as lead magnesium niobate (PMN), and liquid crystal dispersed polymers.

The proper arrangement of the prisms depends on the material of which the prisms are made. For example, for the prisms <b>60</b>, <b>62</b>, <b>64</b>, <b>66</b> shown in FIGS. <b>4</b>A(i) and <b>4</b>A(ii), materials with linear E-O properties are used, such as, for example, single crystals including lithium tantalate, and lithium niobate, or any other suitable single crystal material. In this configuration, the prisms <b>60</b> and <b>62</b> have the same domain that is opposite the domain of prisms <b>64</b> and <b>66</b>. To generate the desired electric field across one or more of these prisms, the appropriate voltages are applied to the respective electrodes <b>60</b><i>a </i>and <b>60</b><i>b</i>, <b>62</b><i>a </i>and <b>62</b><i>b</i>, <b>64</b><i>a </i>and <b>64</b><i>b</i>, and <b>66</b><i>a </i>and <b>66</b><i>b</i>, of the prisms <b>60</b>, <b>62</b>, <b>64</b>, and <b>66</b>. Each of the electrodes is typically a layer of any suitable conductive material such as, for example, gold, nickle, or titanium.

By applying voltages with opposite polarities to the prisms with the same domain, that is to prisms <b>60</b> and <b>62</b> or to prisms <b>64</b> and <b>66</b>, or voltages with the same polarity to prisms with opposite domains, the angles that the light beam (<b>38</b>′ or <b>40</b>) is deflected by the prisms is additive. For example, if a voltage V<b>1</b> is applied to the electrodes <b>65</b><i>a </i>and <b>65</b><i>b </i>of the prism <b>64</b>, the beam is directed along a path B<b>1</b>, and when a voltage V<b>2</b> having the same polarity as V<b>1</b> is applied to the electrodes <b>61</b><i>a </i>and <b>61</b><i>b </i>across the prism <b>60</b>, the beam is further deflected such that it is directed along a path B<b>2</b>.

Referring now to FIG. 4B, there is shown a deflector <b>16</b> with a configuration of prisms <b>70</b>, <b>72</b>, <b>74</b>, <b>76</b> that uses a “domain inversion” process. Such a “domain inversion” prism is produced by subjecting a single crystal to a high voltage under high temperatures during the fabrication process. Depending on the type of crystal, the high voltage applied to the crystal is approximately in the range of 300 volts to 12,000 volts, and the temperature is in the range of about 200° C. to 1200° C.

Like the prisms of FIG. 4A, each of these prisms <b>70</b>, <b>72</b>, <b>74</b>, and <b>76</b> also includes a pair of electrodes to which a voltage is applied to create an electric field across the respective prism. With these crystals, the domain of the prisms <b>70</b> and <b>72</b> is the same, and the prisms <b>74</b> and <b>76</b> have an opposite domain. When the voltages V<b>1</b> and V<b>2</b> having the same polarity are applied to the prisms <b>70</b> and <b>74</b>, respectively, the light beam will be directed along a path B<b>2</b> which has a deflection angle that is double that of the path B<b>1</b> along which the light beam is directed when only the voltage V<b>1</b> is applied.

Shown in FIG. 4C is an arrangement of prisms <b>80</b>, <b>82</b>, <b>84</b> of the E-O deflector <b>16</b> which uses materials with quadratic E-O properties, such as E-O ceramics and polymers. With this arrangement, the beam deflection induced by each prism is additive when the voltages V<b>1</b>, V<b>2</b>, V<b>3</b> having the same polarity are applied to the electrodes (similar to those described above) of the respective prisms <b>80</b>, <b>82</b>, <b>84</b>.

The multi-stage construction of the E-O deflectors shown in FIGS. 4A-4C enable the use of simple control schemes such as digital controlling. In such schemes, a simple voltage power supply with “on-off” logic circuits is suitable for the control process. A typically digital control scheme for the embodiment of the deflector <b>16</b> illustrated in FIG. 4C is shown below in Table 2. In this scheme, when no voltage is applied across any of the prisms <b>80</b>, <b>82</b>, <b>84</b>, the light beam remains undeflected and is transmitted along the path B<b>0</b> to the output fiber <b>90</b>. When a voltage V<b>1</b> is applied to the prism <b>80</b> the beam is directed along the path B<b>1</b> to the output fiber <b>92</b>. And when the voltage V<b>2</b> (having the same voltage and polarity as V<b>1</b>) is applied to the prism <b>82</b>, the beam is directed along a path B<b>2</b> to the output fiber <b>94</b>, and similarly when the voltage V<b>3</b> is applied to the prism <b>84</b>, the beam is directed along a path B<b>3</b> to the output fiber <b>96</b>. Thus, by applying a voltage Vm across one of or a combination of the three prisms <b>80</b>, <b>82</b>, or <b>84</b>, an incoming light beam can be switched to any one of the output fibers <b>92</b>, <b>94</b>, or <b>96</b>. And of course, if no voltage is applied across the prisms, the light beam is transmitted along the path B<b>0</b> to the output fiber <b>90</b>.

<tables><table frame="none" colsep="0" rowsep="0"><tgroup align="left" colsep="0" rowsep="0" cols="1"><colspec colname="1" colwidth="217pt" align="center" /><thead><row><entry namest="1" nameend="1" rowsep="1">TABLE 2</entry></row></thead><tbody valign="top"><row><entry namest="1" nameend="1" align="center" rowsep="1" /></row><row><entry>Digital Control Scheme</entry></row></tbody></tgroup><tgroup align="left" colsep="0" rowsep="0" cols="3"><colspec colname="offset" colwidth="14pt" align="left" /><colspec colname="1" colwidth="84pt" align="center" /><colspec colname="2" colwidth="119pt" align="center" /><tbody valign="top"><row>&lt;entry /&gt;<entry>Prism Status</entry>&lt;entry /&gt;</row></tbody></tgroup><tgroup align="left" colsep="0" rowsep="0" cols="5"><colspec colname="offset" colwidth="14pt" align="left" /><colspec colname="1" colwidth="14pt" align="center" /><colspec colname="2" colwidth="56pt" align="center" /><colspec colname="3" colwidth="14pt" align="center" /><colspec colname="4" colwidth="119pt" align="center" /><tbody valign="top"><row>&lt;entry /&gt;<entry>V1</entry><entry>V2</entry><entry>V3</entry><entry>Light being switched to</entry></row><row>&lt;entry /&gt;<entry namest="offset" nameend="4" align="center" rowsep="1" /></row><row>&lt;entry /&gt;<entry>Off</entry><entry>Off</entry><entry>Off</entry><entry>Path B0 → Fiber 90</entry></row><row>&lt;entry /&gt;<entry>On</entry><entry>Off</entry><entry>Off</entry><entry>Path B1 → Fiber 92</entry></row><row>&lt;entry /&gt;<entry>On</entry><entry>On</entry><entry>Off</entry><entry>Path B2 → Fiber 94</entry></row><row>&lt;entry /&gt;<entry>On</entry><entry>On</entry><entry>On</entry><entry>Path B3 → Fiber 96</entry></row><row>&lt;entry /&gt;<entry namest="offset" nameend="4" align="center" rowsep="1" /></row></tbody></tgroup></table></tables>

Note that the amount the light beam has to be deflected so that it is properly directed to one of the output fibers is determined by the optical properties of the second collimating lens <b>24</b> and the spacing between the output fibers <b>18</b>, <b>20</b>, and <b>22</b> for the array shown in FIG. 1, or the output fibers <b>90</b>, <b>92</b>, <b>94</b>, and <b>96</b> for the fiber array shown in FIG. <b>4</b>C. In general, a more compact optical switch <b>10</b> requires less power to operate. For such compact devices, the spacing between the output fibers should be as small as possible, but large enough to satisfy cross-talk specifications.

As an example, a SLW-1.8 graded index (or GRIN) lens such as that fabricated by NSG, Inc., Japan, can be employed as the second collimating lens <b>24</b> for the output fibers <b>90</b>, <b>92</b>, <b>94</b>, and <b>96</b> of the four-fiber array shown in FIG. <b>4</b>C. With such an array, the deflection angle required to redirect the light beam from one fiber to an adjacent fiber is about 0.89° if the spacing between the fibers <b>90</b>, <b>92</b>, <b>94</b>, and <b>96</b> is approximately 30 μm.

However, because a single mode fiber has a cladding size of about 125 μm, special measures have to be taken for reducing the spacing to 30 μm. One way to achieve this is to position a fiber pigtailed channel wave-guide array <b>500</b> illustrated in FIG. 5 between the collimating lens <b>24</b> and the output fibers <b>90</b>, <b>92</b>, <b>94</b>, and <b>96</b> to guide the optical signal from the collimating lens <b>24</b> to one of the output fibers. The array <b>500</b> can be made, for example, by silica-on-silicon methods and includes a set of narrow channels <b>501</b> that are connected to one end a set of wider channels <b>504</b> with four transition channels <b>502</b>. The other end of the wide channels <b>504</b> are connected to the output fibers <b>90</b>, <b>92</b>, <b>94</b>, and <b>96</b> discussed above.

The variation in the refractive index, Δn, required to deflect a light beam by an angle, α, of 0.89°, is obtained from the relationship

<maths><formula-text>Δ<i>n=αd/L</i></formula-text></maths>

where d is diameter of the beam, and L is the effective length of a single prism. For d=0.3 mm and L=6 mm, Δn=0.000775. Shown in Table 3 are the estimated voltages that have to be applied to three illustrative materials to produce this required variation in the index of refraction, as well as the estimated switching times achieved with deflectors made with these materials.

<tables><table frame="none" colsep="0" rowsep="0"><tgroup align="left" colsep="0" rowsep="0" cols="1"><colspec colname="1" colwidth="217pt" align="center" /><thead><row><entry namest="1" nameend="1" rowsep="1">TABLE 3</entry></row></thead><tbody valign="top"><row><entry namest="1" nameend="1" align="center" rowsep="1" /></row><row><entry>Voltage Requirements and Switching Times</entry></row></tbody></tgroup><tgroup align="left" colsep="0" rowsep="0" cols="4"><colspec colname="1" colwidth="77pt" align="left" /><colspec colname="2" colwidth="42pt" align="center" /><colspec colname="3" colwidth="28pt" align="center" /><colspec colname="4" colwidth="70pt" align="center" /><tbody valign="top"><row>&lt;entry /&gt;&lt;entry /&gt;&lt;entry /&gt;<entry>Liquid crystal</entry></row><row><entry>Material</entry><entry>LiTaO</entry><entry>Ceramic</entry><entry>polymer</entry></row><row><entry namest="1" nameend="4" align="center" rowsep="1" /></row><row><entry>Estimated Voltage (V)</entry><entry>1000</entry><entry>100</entry><entry>10</entry></row><row><entry>Estimated switching time</entry><entry>˜nsec</entry><entry>˜μsec</entry><entry>˜msec</entry></row><row><entry namest="1" nameend="4" align="center" rowsep="1" /></row></tbody></tgroup></table></tables>

Accordingly, when the deflector <b>16</b> is made from materials such as LiTaO, the optical switch <b>10</b> is capable of switching times in the order of nanoseconds. Note that regardless what material is used, the E-O deflector <b>16</b> requires very low currents to operate. Thus, when in use, the power consumption of the optical switch <b>10</b> is typically less than one watt.

While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.